You're probably getting used to it, but we're going to get philosophical again. Just because it is so easy to start a paid search program doesn't mean that you should not do some thinking first. In fact, because it is so easy to shoot from the hip, it is probable that some of your competitors run their programs that way. So, just a little thinking on your part can pay big dividends.
We spend the rest of this chapter explaining how to optimize your paid search campaign, but before we do, do you know what it is worth to you? As we have throughout this book, we harp on measurements. So, let's look first at what we know about the value of paid search.
Look for Value
Gunning for #1. That is the philosophy of the rookie search marketer. But let's take a step back and see whether we can adopt the philosophy of a cagey veteran marketer instead. The savvy marketer is looking for value, so let's examine how you find value in paid search.
We know that paid search drives traffic. You learned earlier that shopping search, in particular, drives visitors that are much more likely to buy than garden-variety searchers. But what about paid placement? Atlas DMT, a paid search software vendor, has demonstrated that paid placement can be highly effective in driving traffic, and that garnering the top positions drives the most. (Not exactly rocket science, huh?)
What might surprise you is the disparity in traffic from each paid placement rankthe #1 position drives about ten times more traffic than #10. With Google, even the drop-off from #1 to #2 is dramaticabout 40 percent. Although Atlas studied only paid placement, you can assume that having the top ranking in a shopping search category is valuable, too.
So what's wrong with gunning for #1? It seems to make a lot of sense. You are in paid search to drive traffic and ranking #1 drives the most traffic. Well, if you have an information site, you might be happy with merely generating trafficperhaps your goal is to sell advertising on your site. For most of you, however, Web conversions are your focus. So, does getting a top paid search ranking give you the greatest number of conversions? Not always. Another study from those busy Atlas folks showed that the highest conversion rate is generally found for the top ranking results, but conversion rates vary by keyword demand. The 20 percent of keywords in the greatest demand had high conversion rates for the highest-ranked results, but the remaining 80 percent were all over the mapthose 80 percent lower-demand keywords frequently produced higher conversion rates for their lower results than for top slots. Some Google results that landed at positions 810 had higher conversions than the #1 for the same keyword, for example. Less markedly, less-popular Yahoo! queries often showed similar conversion rates for the #1 and #10 results.
So what do we make of this research? Well, the good news is that you can pay considerably less than what it costs to rank #1 for many keywords and still convert at a high rate. However, the bad news is that you will have to redouble your keyword research efforts to find those low-volume keywords. You will also have to work harder at measurements, so that you can see when you are converting and when you are not. We also might conclude that Google's ranking algorithm might give the savvy search marketer an edge, because it places so much value on clickthrough. Perhaps those #8 ranking keywords are converting well in part because they get high clickthrough, and their per-click costs are modest.
Search marketers seem to be warming to this approach of purchasing deeper keywords. In 2004, the average advertiser's number of keywords rose 50 percent, reported BusinessWeek magazine.
No studies yet tell us whether shopping search conversion rates are high for lower-ranked results, but because most shopping engines charge every competitor in the product category the same per-click fee, there is no reason to be anyplace except at the top. Ranking lower usually does not save you money the way it can for paid placement.
You also need to develop your philosophy on local search and geographic targeting. If your business is listed in the printed phone book, you might as well take advantage of the free listings by Internet Yellow Pages providers. But should you employ enhanced listings for a $10 monthly fee? Should your paid placement campaigns use geographic targeting? You should think through the advantages of such programs and decide whether you want to test their effectiveness for your site or skip them completely.
So the first part of our philosophy is that you are looking for value, regardless of the rankings. Although top rankings generally drive more traffic to your site, you saw that rankings do not always correlate with traffic. And besides, your goal is to drive the most conversions at the lowest pricethat is what "value" means. We know that we might need to do a little extra work to find value, but the studies say that the value is there.
Play the Market
Paid search is a marketplace, so you might as well apply what you know about markets as part of your paid search philosophy. All markets are made up of buyers and sellers, and prices fluctuate based on supply and demand. When supply is high relative to demand, prices are cheap. As more and more buyers chase after fewer and fewer goods, however, prices rise, the economists will tell you. So how does that apply to paid search? There aren't any "goods," after allit's just a spot on a Web page, isn't it?
Let's take a tour of the important market principles that underlie paid search markets, starting with the basics.
Paid Search Market Basics
Although economists like to talk about "goods," market forces apply to services, too. Think about how passenger seat prices are set for an airline. Tickets bought far in advance are usually discounted because the airline collects the passenger's money long before the flight. Nonrefundable tickets are cheaper than refundable tickets because the airline locks in a sale. Because early payments and guaranteed sales have value to the airline, the airline lowers the price of the ticket in those circumstances. But what happens as the day of the flight approaches? If there are few remaining seats, the airline raises prices, so that last-minute business fliers pay top dollar because they have no alternative. If many seats remain, however, the airline might deeply discount each ticket hoping to attract impulse vacationers, because every ticket sold, even at a deep discount, is worth more than an empty seat.
Paid search markets have some of the elements of airline pricing. Shopping search engines do not charge the same per-click fee for every category because some categories are worth more than others to search marketers. Categories for products with higher prices (and presumably higher profits) tend to cost more than those for inexpensive items, just the way airlines try to offer vacationers bigger discounts than business travelers.
Other forms of advertising follow market principles, too. Commercials on highly rated prime-time TV shows cost more than those during the Late Late Show. When more people watch the show, advertisers pay more to reach them. For the same reason, high-demand keywords tend to cost more than less-popular ones because search marketers are paying for traffic (and they hope conversions).
But shopping search and paid placement are two different kinds of markets. You do not typically bid on your per-click fees with shopping search engines, where the seller sets the price and the buyer decides whether to pay the price or not. Paid placement is an auction market, in which only minimum bids are set and the price is set by whatever buyers are willing to pay.
For shopping search, your philosophy, as you saw earlier, is to find value. You need to determine which products you should sell through shopping search and which ones do not pay. You must know when the per-click fee charged for your product category by a shopping search engine does not allow you to turn a profit, so that you can walk away from that "opportunity." Later in the chapter, we show you how to calculate that. If the price is too high for you, you might check other shopping search engines to see whether they are cheaper or avoid them altogether.
Auction markets have a different ethos than fixed-price markets. When the number of buyers is high, the prices at auction markets tend to be high, too, because more buyers are chasing fewer goods (and perhaps because the prices more accurately reflect the value of what is being sold). Auction markets also result in more sales, because items that would go wanting if priced too high are sold at a price that might disappoint the seller, but they are still sold.
Think about how eBay works. Many unique items sold on eBay would have brought a fraction of their price at a garage sale, simply because a higher number of serious collectors are exposed to the item on eBay. On the other hand, commonplace items sell for cheap prices because of the high number of sellers who make them available compared to the size of the buyer market for those items.
Economists call this phenomenon scarcityscarce items cost more when many people desire them. And scarce items fetch far higher prices at auction, which is why auction houses are always selling memorabilia or a celebrity's estate. If there is only one of something, and two rich people both want it, the price can go sky-high.
Paid placement follows the scarcity principleonly one result ranks #1 for every keyword within a search engine. If two bidders are determined to be #1 for a keyword, prices can skyrocket in a matter of hours.
Efficiency Is Everything
When you are faced with markets such as shopping search and paid placement, you can watch prices gradually rise to the highest sustainable point. As more and more search marketers begin to play the paid placement game, prices tend to rise. They aren't making any more #1 positions for the digital cameras keyword. There's only one. The more bidders that want it, the higher per-click fees will go. (Similarly, the more players enter the digital camera product category, the higher the per-click fees that the shopping search engines can charge.)
Economists call this tendency market efficiencysellers will, over time, maximize the price based on the value to the buyers. And economic price theory holds that, over time, all benefits of an item will be incorporated into its price as buyers' knowledge becomes "perfect." Well, nothing is perfect in real life, only in economic theory, so inefficiencies exist everywhere in markets, just waiting to be exploited.
Your job: Find them. Think about how the stock market works. If everyone knew exactly what every stock was worth at any given moment in time, there would not be much point to buying and selling. However, one of the reasons that shares of stock are bought and sold is that two people have very different opinions about what the stock is worth. One reason this is so is that, economic theory notwithstanding, some of the "knowledge" that people have is irrational. It makes no sense at all. Search marketing is no different. If you are rational, you will usually outdo emotional types and guessers. It is like the stock market in that you need information to get an edge, but not hot tips. The information you need is right under your nose and you can gather it.
Unlike the stock market, paid search is not worth the same amount to every company. Even if every site paid the same 25¢ per click, some companies would make more money on that click than others. If you gather the information available to you, you will know what each keyword's position is worth to you, so you will never overpay.
But it is usually more complicated than that, because the shopping search engines might be charging more than you can afford for some products, or the high rankings in paid placement are bid through the roof. How can others afford to pay when you cannot?
Two answers are possible. One is that they are nuts. They are bidding to keep that #1 position, and they lose money on every sale. If that's the case, you probably have a short-term problem; economics has this funny way of correcting problems such as money-losing bidding. Although annoying, you can withstand that situation for as long as it lasts.
But you might, in fact, have a more serious problem. Perhaps they can pay higher prices because they are more efficient than you are. Maybe they can make money at that high rate, but you cannot. Several reasons could account for this:
So, to take advantage of inefficiencies in the market, you must make your business as efficient as possible, so that you will have all of those advantages over your competitors, rather than the other way around. As a search marketer, you will not have much impact on whether your company is the low-cost producer in your field, and you will not have a great deal to do with how fast your inventory turns, but there are a lot of things you can affect, which we discuss later in this chapter.
Keep in mind that the only way to affect anything is to have the necessary information. As you drive for efficiency, you want to know the conversion rate for every product you sell (for shopping search) and for every keyword you buy (for paid placement), but you want to know much more. Are conversions higher
If so, you can take advantage, as we explain later. By gathering the most detailed information possible, you will know the right situations to play the market and when to hang back. But ratcheting up your efficiency is only part of the game. You also need to learn about an old business maxim of using other people's money.
Other People's Money
Some of the richest people around got there by investing other people's money. If the investment fails, they lose nothing; if it succeeds, however, they will get something out of it. Frequently these smart people get a management fee or a cut of the proceeds for their contribution to the success.
Smart search marketers can use other people's money, too. Here are several ways to do that:
Any of these techniques can make your budget go much further than if your philosophy is to go it alone, but you will be forced to share more information with your new-found partners than ever before, so you need to have close relationships for it to work well.
So far, you have learned a lot about how to develop a paid search philosophy, but the most important lesson is humilityyou will never get it completely right. Next we show you how to get it a little less wrong every day.
Iterate, Iterate, and Then Iterate Some More
Here's a secret: Paid placement won't work. At least, it won't work at first. When you start, you will find that your ads get so few clicks that they are disabled by the paid placement engine. Or you will find that your ads get loads of clicks, but almost no sales. Expect this. Expect it not to work. Your program will start badly.
But no matter how badly you start out, if you are willing to track your numbers and keep tuning your program, you will eventually succeed. The secret is to start small. You know you are not going to get it right, so why not lose as little as possible while you learn? You learn just as much about horse racing betting at the $2 window as you do at the $100 windowyou just have a lot more money left over after the last race.
The secret is not how well your paid search campaign startsit is how you improve it every day. If you have chosen the right metrics to track, and you religiously measure your success, you can keep adjusting until you are as efficient as you can be.
If you can afford to start out with a large budget, the best way to spend it is on as broad a set of targets as possible. Don't shoot your wad of cash trying to be #1 for a popular query. Instead, buy as many keywords as you can, even if they are very low positions. Buy different keywords every day if you need to spread things around. If it doesn't cost any more to set up, put your entire catalog into a shopping search engine. Why? Because that way you will gather the most data. You will start getting click rate data and conversion data for as many targets as possibleas many paid placement keywords and as many shopping search products as you can afford. Do not overspend. Spend as little as possible on the widest possible set of targets.
After you have collected this early data, see who the winners are. Do you have some keywords that are highly profitable? Spend more on them to see whether higher rankings bring more traffic (and more profits). Do you have catalog items that are not worth the per-click costs for shopping search? Watch them carefully and remove them if they do not improve. In the rest of this chapter, we show you how to improve your clickthrough rates and your conversions rates so that you make more and more of your targets profitable. But that is not the most important lesson.
The most important thing to learn is that you must iterate. You will not get it right the first timedon't try. Instead, commit yourself to a constant feedback loop. Then adjust to that feedback for continuous improvement.
Let's look at how Snap Electronics handled the situation. They budgeted $150,000 for paid search in the first year (as discussed in Chapter 8), which comes to about $12,500 a month. They knew that they needed to test, so they decided that they would spend much lessabout $1,000to start, believing that they could experiment to see what works before putting a lot of money behind their campaign later. Snap bought dozens of keywords for digital cameras and ran their keywords for just a few hours eachthe same hours each dayfor a week. By analyzing their results, they had a much better idea of the keywords that would be the most profitable in a full-blown campaign.
Start out by going broad to gather the most information possible, but do it at the lowest possible cost. Then study your results to optimize your success. Like the stock market, past performance is no guarantee of the future, so be alert to ongoing results so that you can adjust your valuations.
Clearly, you have a lot to think about before plunging into your first campaign, but don't feel like you have to figure out everything before you start. One of the biggest advantages of paid search is its incredible adjustability. No matter what you start out doing, you can see what works and what does not in a few days and then do something else. The most important philosophy for paid search is no different than for search marketing as a wholetry something, measure its success, and then try something better. That way, you optimize your campaigns for the biggest return. Let's go step-by-step through a paid search campaign to see how it's done.